Although previous studies have highlighted advantages of adopting climate-smart agriculture (CSA) practices, their actual capacity to generate long-term and inclusive effects remains unclear. Beneficiaries report positive effects on their living conditions, yet many continue to face limited access to productive resources. This study assesses the socioeconomic impacts of CSA practices as perceived by beneficiaries, characterizes the profiles of the actors reporting the main impacts, and identifies the factors influencing these perceptions. A mixed-methods approach combining a quantitative survey (N = 200) and qualitative interviews with 18 of these respondents was applied to 200 producers benefiting from CSA initiatives in northern Benin, primarily in the communes of N’Dali and Tchaourou. A convergent parallel mixed-methods design was applied: the two strands were collected concurrently and merged at the interpretation stage through a joint display. Quantitative data were analyzed using cross-tabulations with chi-square tests, Spearman rank correlations and binary logistic regressions reporting odds ratios with 95% confidence intervals; the interviews were analyzed thematically. Positive impacts on income are reported by 70.5% of beneficiaries, while, among those reporting improved income, most consider their access to land (58.9%) and water (67.4%) unchanged. The findings reveal three paradoxes that characterize a situation of partial emancipation. First, women, who report gains in proportions similar to those of men, perceive economic gains associated with CSA practices without necessarily achieving structural autonomy or land tenure security: reported autonomy coexists with access to land and water that beneficiaries describe as unchanged. Second, the study highlights a form of economic mobility without security of productive resources (land, water, seeds, and credit). Third, social recognition is not accompanied by a proportional strengthening of real economic power. Perceived impacts are most strongly associated with technical training, age and access to seeds and financing, rather than with land or water security. Because the design is cross-sectional and based on self-reported perceptions, the findings identify associations and perceived experiences rather than causal effects of the interventions. These observations call into question the effectiveness of current gender-sensitive approaches in CSA-focused projects.